IPFS Weekly Call, 5 Sep 2016

Previous Meeting Next Meeting

⏯

youtube image

►

From YouTube: IPFS go-ipfs discussion

Description

See https://github.com/ipfs/pm/issues/162

A

Hello welcome to Goa PFS. This is the first of three discussion calls that will do it. It's a big part of my PFS that we have on Monday the lead, for this is Jeremy aka. Why are you sleeping and there are notes which are on the link in the sprint issue and I- think you've also been posted to IRC. At this point, I was.

B

A

C

Fine, how many kuba.

A

Kuba, are you taking yeah.

D

B

A

Birthday boy I will be taking notes, sweet Jeremy go ahead.

C

All right, cool, so I think the new proposed format for these types of calls is just to go over our milestones and our progress on each of them. So that's what I'm going to do. First, we have the four three release candidate, for we have all the PRS for this thing merged, which is cool.

C

We made some good progress on that we fixed a so there is a race condition that got fixed in RC Jeremy.

B

Could you post a link.

C

To what page you're looking at just so, everyone can follow along just yeah I'm, looking I'm like reading from the Internet sorry I'm, just kind of like yeah yeah, so anyways yeah, the release is getting close.

C

We have the running on a couple gateway notes, they're, happy they're, actually using the least number of go routines and on average they use my slower rant than the rest of the things which is SuperDuper cool, fixed a bit swap issue, which means that your want lists aren't going to build up tons of keys a few canceled requests now which they have and yeah so we're just working on a change log or this lease candidate. Then we're going to push that out.

C

Rc4, it's gonna, get some happy fun time and then we're gonna, merge it and officially release oh for three, so yeah next up is the Jay s. Ip FS interrupt me and D hash here have been working on getting things talking to each other and it's we're making it farther and farther. The current bottleneck. Milestone thing we're working on is that JavaScript and cryptography aren't very fun together, so we're trying to figure out some thick curve nonsense.

E

C

E

C

E

Other words? Yes, yes, yes, yes,.

C

E

Go awesome, yeah.

C

It takes 24 seconds to run one little test.

E

C

Yep, so we're getting getting close there. Our goal is to have that, like today, working be cool. Yeah next milestone is our code coverage milestone.

C

There are little progress if any on that this week, because we've been busy with other stuff and adding more tests, kind of got pushed back to later and next, for the main reason we're actually in New York right now is the IPO, the integration.

C

So we have go see IDs written a PR, that's integrating it into gorath EFS, which it'll be integrating the version 0, which is basically exactly what we already have, but using content IDs to represent that information internally, as opposed to just the key stuff. So what will happen is once I get this in nothing will change. You won't notice anything different all like externally. There should be zero, behavior change and all that you'll notice is.

C

If you look at the code, there will be different packages used to represent block locks and objects and stuff, and the cool thing this will. Let us do is when we're ready to switch to the we can just start ingesting the new keys and switch over the version from zero to one on creation and it'll just handle it. It'll be just just file, be fun and wonderful, and so yeah that would be really really cool. um Yeah here here, I was going to get that link. Wasn't nice I'll get that so yeah?

C

That is the extent of our milestones and the progress is week. Kuba you had some other stuff. You were working on which you like to share.

D

Your I was I, was I finally finished working on the matrix interface, alright or as I showed recent does just a few months ago, like it got pushed so I started working on it like one a half months ago before the ten-week, so it could push back a lot but I finally finish it and I think it should be good to go to start implementing other metrics using it I used to implement few metrics in the boxer, but yeah. So that's this I've worked on some other things. Also that cool.

C

That's good I'm gonna be happy to get better metrics in it'll. Make things a lot nicer, especially since our current largest consumer of memory in co-op EFS on average, is our metrics.

D

B

D

We are using two metrics libraries. We have one massive library that is in. Go that a store yes and this one. It is consuming a lot of memory and we have and the primitives isn't cool I.

B

Can give a very, very short update about the core API milestone? Yes, so none of the tasks- and it has really been finished in but I'm working on the first one, which is to have the Gateway code and go ipfs, use a very minimal core API, so that entails like bringing some kind of core API and to go a PFS and implementing towards of the object API so that the gateway can use it. That's it I'm. It's a work in progress right now,.

C

C

That'll be really exciting to be able to keep breaking things out more because, like we just we're getting really close timing, the DHT code broken out into its own package, which once that's done. The IPS hub tool will stop breaking all the time because we'll be able to actually make it.

B

Yeah, you know.

C

Nice mended code it'll be cool.

B

That tool is such.

C

B

Build right now, basically,.

C

Only builds like the day we fix it and.

C

Depressingly enough, it'll be better safe, so yeah that is copyof s and our progress on milestones. Any questions.

C

That looks like a No, so I think we're done. That was 12 minutes good job.

F

Well, do you see you gave an update on the math out students, you want to talk about butter, or some means that you would like people to work on for next week or her this week.

C

Things I would like people to work on, so we have a lot of issues that are tagged as both Help Wanted and verify either of those are things that, like you, can go, grab a thing and help out. So if it's take Help Wanted, that's gonna be like stuff, that's ready to be implemented should be well defined and you should be able to just step in and start hacking on something the verify ones are really helpful, because what that entails is hey. We head up a bug.

C

It's it may be been a while since we've encountered it or you know, had to deal with it and we're not sure if it's been resolved and not so, we need somebody to go verify like. Is this been fixed? Is this thing, and so having that done is really helpful because it gives us better confidence about. You know saying: okay: this has been solved as opposed to saying: oh, we think this has been solved, and so anybody who wants to help out getting those would be super cool.

C

It's just a little little time-consuming because you know nature of bugs so yeah. That's that do you have any input, I, think everything everything is great cool.

C

Alright anybody else or are we calling it I.

A

Think that's it or calling it cools everyone. There will be another chat for JSF BFS in 15 minutes at UTC, 17 34. Now. Thank you very much and if you have any questions or anything feel free to ping us on IRC or.